miRNA Analysis by Next-generation Sequencing and its Prognostic Importance in Non-small Cell Lung Cancer

Abstract

Lung cancer is the leading cause of cancer death worldwide, with an estimated 1.8 million new cases diagnosed each year. Non-small cell lung cancer (NSCLC) is the predominant histological type, accounting for approximately 85% of all lung cancers. Despite significant advances in smoking cessation, early detection and genetic understanding of the disease, the 5-year overall survival has remained at ~15% over the past few decades. Post-surgical chemotherapy has been shown to improve the survival of some early-stage NSCLC patients, but the overall benefit is modest. To optimize the toxicity/benefit ratio, better diagnostic tools are needed to identify low-risk individuals who can be spared from unnecessary intervention, while avoiding undertreating high-risk patients. Over the past decade, a multitude of prognostic gene signatures has been published, however, none are currently used in the clinical setting to guide therapeutic decision-making. It is hypothesized that the integration of multidimensional high-throughput molecular data may better capture the molecular heterogeneity of lung cancer patients, resulting in more robust biomarkers for patient classification. To aid in the prediction of patient outcome using multidimensional molecular data, I have worked on different stages of the process of biomarker development with respect to microRNAs (miRNA) and their potential added prognostic value to gene signatures. First, I conducted a comprehensive comparison of miRNA profiling technologies, including next-generation sequencing, a microarray platform and the NanoString nCounter System. Next, data preprocessing methods were evaluated to develop a standardized pipeline for the preprocessing of small RNA-seq data, with the end-goal of retrieving miRNA count abundance profiles. Finally, miRNA profiling was performed on tumour specimen resected from early-stage NSCLC patients and combined with gene expression profiling data for the development of an integrated miRNA-mRNA risk classifier for predicting prognosis. The knowledge gained from these studies provides technical insights for the analysis of miRNAs and biological insights for the development of more robust prognostic classifiers.